Effect of Pore Width on Micropore Filling Mechanism of SO2 in Carbon

Effect of Pore Width on Micropore Filling Mechanism of SO2 in Carbon Micropores ... View: PDF | PDF w/ Links | Full Text HTML ... Activated Semi-coke ...
6 downloads 0 Views 74KB Size
J. Phys. Chem. B 1998, 102, 2863-2868

2863

Effect of Pore Width on Micropore Filling Mechanism of SO2 in Carbon Micropores Zheng-Ming Wang and Katsumi Kaneko* Physical Chemistry, Material Science, Graduate School of Natural Science and Technology, Chiba UniVersity, Yayoi 1-33, Inage-ku, Chiba 263, Japan ReceiVed: October 8, 1997; In Final Form: February 11, 1998

The mechanism of SO2 micropore filling in slit-shaped micropores was examined by direct calorimetry and potential calculation, and the effect of the pore width on the SO2 micropore filling mechanism is discussed. Four kinds of activated carbon fibers of different pore widths have been used, and both granular activated carbon and microporous Boehmite aggregates were used for comparison. The differential energy of SO2 adsorption (qd) on ACFs varied with change of the average pore width. An increase of the micropore width leads to a decrease of qd in SO2 micropore filling by decreasing the enhanced micropore field. Molecular potential calculations gave good agreement with the experimental adsorption energy when the image potential due to the SO2 dipole moment was considered. The image potential contributes 25-32% to the total molecular potential, indicating its importance in micropore filling of dipolar SO2.

Introduction

Experimental Section

Micropore filling has been studied extensively over the last 30 years.1-9 However, the relationship between the mechanism of micropore filling and molecular states upon micropore filling is not sufficiently elucidated. Direct calorimetry can provide useful information on the states of molecules adsorbed.10-12 Hence, there has been active calorimetric research on micropore filling recently.13-26 Also, newly developed computer simulation studies need calorimetric data on micropore filling of various kinds of molecules.27-30 Specific interaction should play an important role in the micropore filling of a polar molecule. There are many important polar molecules having large dipole moments. H2O is a representative dipolar molecule. Recent in-situ X-ray diffraction studies showed that water molecules form an ordered structure in the carbon micropore, indicating that the contribution of hydrogen bonding is significant in the intermolecular interaction.31,32 SO2 has a large dipole moment and the adsorption on activated carbons has been studied with relevance to atmospheric environmental problems. As the intermolecular interaction of SO2 has no contribution from hydrogen bonding, we can understand how the molecular dipole is associated with the micropore filling. In particular, activated carbon fibers (ACFs) have controlled uniform micropores,33-39 and the micropore filling mechanism of SO2 on the microporous carbon can be examined as a function of the pore width. The preceding work using calorimetric measurement, and a molecular potential calculation for SO2 adsorbed in the narrow slit pore, where only a single SO2 layer can be filled, showed a dipole-oriented structure of SO2.40 The importance of the dipole-induced dipole interaction in the molecular surface interaction was indicated. Systematic examinations of SO2 adsorption on various ACFs having different pore widths should provide a greater understanding of the micropore filling mechanism of the polar molecule. This study describes the effect of the micropore width on the micropore filling of SO2 using the combined approaches of the calorimetric measurement and molecular potential calculation.

Pitch-based ACFs (denoted as P10, P15, P20, and P25; Osaka Gas Co.), with micropore width increasing with the serial number, were used for this study. The granulated activated carbon (denoted as AC) from Takeda Yakushin Co. and Boehmite (γ-AlOOH) aggregates (denoted as BA) having slitshaped micropores41 were used for comparison. The microporosity of samples was determined by N2 adsorption at 77 K. The SO2 adsorption isotherms were determined gravimetrically at 303 K. The irreversible adsorption amounts of SO2 were determined after evacuation during one night at 303 K. The differential adsorption energy of SO2 on ACFs was measured by a standard twin-type calorimeter (Tokyo Richo Co.), having a high sensitivity of 28 mV/deg. A released heat, dQint, was determined, when dNa molecules of SO2 were adsorbed on a sample. Then, the differential adsorption energy, qd, of SO2 was calculated by the following equation:

qd ) (dQint/dNa)T)303K

(1)

Molecular Potential Calculation Method. The detailed calculation method of the interaction potential of a dipolar SO2 molecule with the graphitic slit pore was described in our preceding work.40 Here, we give a brief description. Physical characterizations of ACFs showed that ACF is composed of nanographites and the micropores can be approximated by the slit-shaped graphite pores.37 Furthermore, ACF exhibits good semiconductivity.42 The polar interaction of a SO2 molecule with the graphitic slit pore, therefore, was calculated with the induced image potential method. The nonpolar interaction of SO2 with the graphitic slit pore was calculated with the aid of the 10-4-3 potential proposed by Steele.43 In the slit-shaped micropore, the nonpolar molecule-pore wall interactions from opposite pore walls are overlapped, as shown in eq 2:

S1089-5647(97)03270-7 CCC: $15.00 © 1998 American Chemical Society Published on Web 03/31/1998

2864 J. Phys. Chem. B, Vol. 102, No. 16, 1998

Wang and Kaneko

Ψh(zm) ) 4πFσsf2∆sf[(1/5)(σsf/(d - zm))10 - (1/2)(σsf/(d - zm))4 σsf4/(6∆(0.61∆ + d - zm)3)] + 4πFσsf2∆sf[(1/5)(σsf/(d + zm))10 - (1/2)(σsf/(d + zm))4 - σsf4/(6∆(0.61∆ + d + zm)3)] (2) where d is the distance between the nuclei of the first layer carbon atoms of the pore walls and the center of the micropore wall adsorbed molecule, and is calculated as the sum of half the pore width and the radius of carbon atom (rc ) 0.071 nm); zm is the distance from the interacting center of an adsorbent molecule to the center of the pore; F is the number of carbon atoms per unit volume of a graphitic lattice; σsf and sf are the distance contact and the potential well depth of a Lennard-Jones (LJ) potential between an adsorbent molecule and a carbon atom, respectively; and ∆ is the intergraphitic layer distance. The σsf and sf were calculated from the LJ parameters of SO2 and a carbon atom by the Lorentz-Berthelot rule.44 The following values of ∆, F, σsf, and sf were used for the calculation: ∆ ) 0.335 nm, F ) 11.36 nm-3, σsf ) 0.385 nm, and sf ) 84 K. The polar molecule-slit pore interaction was calculated by the following equation with the image potential approximation:

Ψmi(zm) ) (-µ2/8)[1/(d - zm - ∆/2)3 + 1/(d + zm - ∆/2)3] (3) Here µ stands for the dipole moment of SO2 (µ ) 1.61 D). The position of the image plane was assumed to be situated above the pore wall by ∆/2 using Crowell’s approximation for adsorption of SO2 on the basal plane of the graphite.45 The axis of the SO2 dipole moment was assumed to be normal to the graphitic wall; that is, the maximum induced image potential was presumed. Summation of eqs 2 and 3 gives the total interaction potential of a SO2 molecule with the graphitic micropore:

Ψt(zm) ) Ψh(zm) + Ψmi(zm)

(4)

From eq 4, the distance of the minimum potential, zm*, and the minimum potential well depth, Ψ(zm*), can be obtained. The total potential well depth Ψ0 of a molecule with a single graphite surface was calculated; enhancement of the adsorption energy in the micropore is expressed by Ψ(zm*)/Ψ0. Results and Discussion Characterization of Samples. All of the N2 adsorption isotherms of ACFs, AC, and BA were typical of type I. The micropore volume, V0(N2), and the specific surface area, as, of ACF and AC were obtained from the Rs plot of the highresolution N2 adsorption isotherm based on the standard data of N2 adsorption on nonporous carbon black;37 that of BA was obtained from the Rs-plot analysis on nonporous oxides.46 The average pore width wm was obtained by geometrical calculation using V0(N2) and as. These pore parameters are shown in Table 1. The pore width of ACF is in the range of 0.79-1.45 nm. The micropore volume and pore width of AC are similar to those of P10. However, AC has a wider pore size distribution than P10, because the adsorption isotherm at a low P/P0 range was more gradual than that of P10. The pore width of BA, whose surface has a hydrophilic nature, is comparable to that of ACF or AC. SO2 Adsorption. Figure 1a shows the SO2 adsorption isotherms on ACF samples. Here the amount of adsorption is

Figure 1. SO2 adsorption isotherms of ACF samples with ordinates of (a) f and (b) φ.

TABLE 1: Microporosity of Samples sample V0(N2)/mL as/m2 g-1 wm/nm

g-1

P10

P15

P20

P25

AC

BA

0.753 1190 0.79

0.767 1545 1.01

0.971 1765 1.13

1.344 1935 1.45

0.514 1250 0.83

0.24 360 1.3

normalized with the aid of the micropore volume from N2 adsorption V0(N2) and the amount of SO2 adsorption V(SO2) using the bulk liquid density (FL(SO2) ) 1434 Kg m-3 at the boiling point) (f ) V(SO2)/V0(N2)). The SO2 adsorption isotherm of P5 in the preceding paper40 is also shown for comparison. This figure shows distinctly the sensitive dependence of the adsorption isotherm on the pore width: the smaller the pore width, the sharper the uptake at a low P/P0 region. Hence SO2 adsorption needs a stronger interaction of a molecule with the pore. Although the adsorption isotherm of P5 belongs to type I, that of P25, having the widest pores, loses completely the nature of type I; rather, it is close to type III. The reason why the isotherm of P5 has the nature of type I was reported in the preceding paper.40 In the case of P5 whose micropores can accept a single layer of SO2, the contribution of a dipoleinduced dipole interaction due to the dipole-oriented single layer structure gives the marked adsorption at a low P/P0 region. As the adsorption isotherm of SO2 on nonporous carbon black belongs to type III, the polar structure of SO2 is not fit for adsorption of SO2 on the graphitic surface. Then SO2 molecules adsorbed in micropores of P25 have no ordered structure of dipoles intensifying the molecule-pore interaction. SO2 molecules adsorbed in micropores of P10, P15, and P20 should lose the ordering of molecular dipoles in this order. The Dubinin-Radushkevich (DR) equation was applied to the SO2 adsorption isotherms. The DR equation is given by eq 5: 1-3

ln V(SO2) ) ln V0(SO2) - (RT/βE0(SO2))2 ln2 (P0/P) (5) Here V0(SO2) is the saturated SO2 adsorption amount, P0 is the saturated vapor pressure of SO2 (P0 ) 4.56 × 105 Pa at 303

Micropore Filling Mechanism of SO2

J. Phys. Chem. B, Vol. 102, No. 16, 1998 2865

Figure 2. DR plots of SO2 adsorption on ACF samples: (a) P5, (b) P10, (c) P15, (d) P20, (e) P25.

K), β is an affinity coefficient, and E0(SO2) is the characteristic adsorption energy. If V(SO2)/V0(SO2) is expressed by φ (Figure 1b), the isosteric heat of adsorption of SO2, qst,φ)1/e at φ ) 1/e is derived from both eq 5 and the Clausius-Clapeyron equation:47

qst,φ)1/e ) βE0(SO2) + ∆HL

(6)

Here ∆HL is the condensation heat of the bulk SO2 gas (∆HL ) 24.9 kJ mol-1 at the boiling point). Figure 2 shows the DR plots of ACF samples. The DR plot of SO2 on P5 has good linearity except for the high P/P0 range. The DR plot of SO2 on P10 is well expressed by two linear regions. Other DR plots for SO2 are not expressed by the linear regions. As the DR plots for N2 on these ACFs had a good linearity due to a well filling process of N2 at 77 K, the nonlinear DR plots of SO2 on P15, P20, and P25 suggest an incomplete filling of pores with SO2 molecules. Accordingly, we cannot determine precisely the V0(SO2) with the DR plot. Table 2 lists V0(SO2 ) and qst,φ)1/e determined from the DR plots in the high relative pressure range, and the SO2 adsorption isotherms using φ as the ordinate are shown in Figure 1b. Table 2 also shows the ratio of V0(SO2 ) to V0(N2 ). Although V0(SO2 ) increases with the pore width for PIT samples, both qst,φ)1/e and V0(SO2)/V0(N2) decreases with the pore width. Hence ACF samples having a larger pore width and a larger pore volume are not fit for SO2 adsorption.

Therefore, SO2 molecules do not form well-ordered dipoleoriented structures in the micropores of P15, P20, and P25. Figure 3 shows the adsorption isotherms of AC and BA in comparison with those of P5 and P25. The adsorption isotherms of SO2 on both AC and BA are similar to that of P5. Although the pore width of BA is 1.3 nm, being close to that of P25, the SO2 adsorption isotherm of BA rather resembles that of P5. This suggests that the interaction of SO2 with the micropore of BA having surface hydroxyls is stronger than that with the graphitic pore. These SO2 adsorption isotherms were analyzed with the DR equation. The qst,φ)1/e of BA is larger than that of P20 or P25, although the micropore widths are close to each other. The surface hydroxyls on the slit pore walls of BA41 should interact more effectively with the polar SO2 molecule. Table 2 shows results for AC and BA. Both of φ and qst,φ)1/e of AC are comparable to those of P5 due to the presence of narrower micropores. The ratios of the irreversible adsorption amounts of SO2, Wirr(SO2), to W0(SO2), are also shown in Table 2. The irreversible SO2 adsorption is caused by the strong interaction of the SO2 dipole with the local surface active sites. Hence, the larger Wirr/W0 for SO2 denotes the higher concentration of the surface functional groups. As the Wirr(SO2)/W0(SO2) of AC is very large, the SO2 micropore filling on AC is affected by large amounts of surface functional groups. Among the ACF samples, only P10 shows a large value of Wirr(SO2)/W0(SO2) in comparison with P15-P25. We must take into account the presence of surface functional groups for P10, but the pore walls of P15-P25 can be presumed to be nonspecific. BA has a value of Wirr(SO2)/W0(SO2) comparable to that of P10, indicating a relatively weak specific interaction between dipolar SO2 and surface functional groups due to defective structure as well as structural hydroxyl groups. Experimental Adsorption Energy of SO2 on Carbonaceous Micropore. Figure 4 shows the changes of differential adsorption energy, qd, of SO2 on ACF samples with φ. The broken line in the figures denotes the condensation heat of bulk SO2 gas at the boiling point. The qd-φ relation of P10 has two plateau regions in the φ range of 0.1 to 0.7. The large qd below φ ) 0.05 should be ascribed to the presence of the surface functional groups. The Wirr(SO2)/W0(SO2) of 0.1 for P10 agrees well with the above result. The two plateau regions just coincide with the two linear regions in the DR plot. The corresponding qd values are 37.0 and 33.0 kJ mol-1. Hence these two plateau values come from the binodal micropore size distribution. The slight increase of qd above φ ) 0.7 may be attributed to the enhanced intermolecular interaction which was distinctly observed in P5. The qd-φ profile of P15 has a flat region below φ ) 0.4 and a gradual decrease at the higher φ region. The qd value at the flat region is 36 kJ mol-1, being slightly smaller than that of P10 at the first flat region. Therefore, the adsorption energy measurement indicates a similar SO2 adsorption state for both P15 and P10 at the low φ region. As the adsorption energy depends on the micropore width, the micropore width of P15 corresponding to φ < 0.4 should be close to that of P10 for the first plateau region. The gradual decrease of qd in the larger φ region of P15 is ascribed to a wider pore size distribution than P10. P15 was prepared by a longer activation than P10, and the pore size distribution of P15 becomes wider. P20 and P25 have similar qd-φ curves. Except for those in the low φ range, the qd value is almost constant (40kJ/mol); this is attributed to the interaction of

2866 J. Phys. Chem. B, Vol. 102, No. 16, 1998

Wang and Kaneko

TABLE 2: SO2 Adsorption Data sample

(P5)

P10

P15

P20

P25

AC

BA

V0(SO2)/mL g-1 qst,φ)1/e/kJ mol-1 V0(SO2)/V0(N2) Wirr(SO2)/W0(SO2)

(0.32) (37.5) (0.95) (